Having assigned concepts to fundamental categories, imposed some more detailed arrangement into arrays, and identified hierarchies, we now have a set of completed facets for the field of animal welfare. It would be quite possible to generate the thesaurus from the vocabulary in its current state, since most of the thesaural relationships are determined by the immediate context of a term in the systematic structure. Equivalence relationships and those of hierarchy are already evident, and many associative relationships can be derived from the facet internal structure without considering how the different facets relate to each other.

Nevertheless, there are some considerable advantages in bringing the systematic display to a more finished state:

Nowadays, when many taxonomies and other vocabulary tools only exist online, it is reasonable to ask whether the idea of order has any relevance. If the animal welfare vocabulary were to be managed digitally, it might be possible never to think about the overall order and structure of the vocabulary. The facets could sit independently in the digital space, indexing terms could be chosen and assigned without thinking about the order of combination, and retrieval would operate on the same basis, that is postcoordinately. The individual items indexed are dealt with both at the indexing and retrieval stage in isolation from each other, and this is perfectly adequate for a high proportion of information-seeking tasks. A number of retrieval systems using faceted techniques in a graphical interface make use of this approach, notably commercial retail sites which allow the customer to select and combine aspects of a product without having to refer to an overall list.

However, most of the ‘terms’ in such systems are very simple and don’ involve anything in the way of overlap between concepts. In the area of documentation, the majority of items are complex, and the concepts dealt with can be compound both in the verbal form of a term and in its conceptual constitution. Having some rules helps us to sort out this complexity and manage the frequent combinations of concepts.